Abstract Tumor microenvironmental heterogeneity is a major driver of treatment resistance and variability in clinical response to therapy. This heterogeneity arises from variations in cellular phenotypes populating the tumor microenvironment (TME), their spatio-temporal localization and expression of surface markers, often associated with specific biological function – stemness, metabolism, proliferation, immune activation and others. Such features are usually studied through ex vivo immunofluorescence and cytometry to identify cellular phenotypes in TME; however, they cannot be applied in vivo for real time TME analysis. Moreover, current in vivo imaging techniques fail to provide real-time high-resolution visualization of the TME. Routinely employed biopsy tissue sampling through fine needle aspirates is limited by its invasive nature and inability to provide a global and dynamic overview of the TME, thus limiting our ability to study spatiotemporal TME dynamics and identify possible features leading to potentially resistant tumor phenotypes. In this study, we demonstrate the use of a hyperspectral fluorescence microendoscope (HFME) to monitor cellular phenotypes in TME with real-time visualization and high video imaging capability (~17 fps). Current fluorescence video microscopy is limited to simultaneous imaging of no more than 2 molecular markers with potential to be expanded to ~4 markers using dichroic mirrors and point detectors; the HFME demonstrated in this study can currently resolve 6 different molecular markers, simultaneously, using a multichannel linear array detector with potential to expand to 10 or more markers. Using a cocktail of near infra-red fluorophore-antibody conjugates targeted against key molecular (surface) markers of different cells in the TME, we are able to capture real-time TME dynamics at cellular resolution in two pre-clinical models; 1) a xenograft orthotopic mouse model of peritoneal carcinomatosis (disseminated metastases within the abdominal cavity) and imaging epidermal growth factor receptor (EGFR), CD44, CA125 (MUC16), transferrin receptor, Thomsen-Friedenreich carbohydrate antigen (T antigen), and CD45, 2) A syngeneic immunocompetent KPC cell line implanted mouse model of pancreatic ductal adenocarcinoma and imaging CD3, CD4, CD8a and CD45. Imaging on these tumor models was performed pre- and post-sub-therapeutic verteporfin (benzoporphyrin derivative monoacid A) photodynamic therapy (PDT, a cytotoxic light-based therapy). PDT treatment resulted in reduction of cancer cell burden, immune cell infiltration and alterations in their relative spatial localization. The results were confirmed by histopathological validation and ex vivo immunofluorescence staining of tumor tissue sections. Collectively, these results demonstrate the capability of HFME to image cancer cell phenotypes and the tumor microenvironment, in real time in live mice. With the ability to monitor cancer growth and treatment effects at a cellular level, HFME can potentially assist in customizing therapies in a patient-specific manner. Citation Format: Mohammad A. Saad, Bryan Q. Spring, Akilan Palanisami, Eric M. Kercher, Ryan T. Lang, Jason Sutin, Zhiming Mai, Tayyaba Hasan. Real-time visualization of tumor cell phenotype and microenvironmental heterogeneity enabled by a hyperspectral fluorescence microendoscope [abstract]. In: Abstracts: AACR Virtual Special Conference: Tumor Immunology and Immunotherapy; 2021 Oct 5-6. Philadelphia (PA): AACR; Cancer Immunol Res 2022;10(1 Suppl):Abstract nr P044.